Why one metric will never measure authority
People want a single authority score, a number that goes up when they are winning. It does not exist, and the third party scores that pretend to be one are blunt approximations at best. Topical authority is a position in the minds of readers and the models of search engines, and positions are described by several signals at once, not summarised in one figure.
The mistake is treating any single metric as the goal. Chase one number and you optimise for that number rather than for actual authority. The discipline is to watch a small set of honest signals together and read their combined direction. When they move up together, you are genuinely gaining ground. When they diverge, the divergence itself is the message.
The signals worth tracking
Four signals, read together, give an honest picture of topical authority. Each one alone is misleading. Together they are hard to fake.
Coverage of the topic
Authority starts with completeness. If you have mapped a topic properly, you know how many distinct questions it contains, and you can measure how many of them you actually answer well. Coverage is the denominator for everything else. A site that ranks for ten queries in a topic of five hundred does not own that topic, however good those ten pages are. This is why the work of mapping a topic completely comes before any attempt to measure authority. You cannot measure your share of a topic you have not mapped.
Ranking spread
Next, how many of the topic's queries do you rank for, and how highly? Ranking spread is more telling than any single keyword position. One page at number one is a page. Forty pages on the first page across a topic is authority. We watch the share of a mapped topic where we appear on page one, because that share rising is the clearest sign the engine is treating us as a topic owner rather than a one query specialist.
Citations and links
How many independent sites reference you, and are they sites that matter in your space? Citation count alone is gameable, so we weight it by independence and relevance. Ten links from real, relevant publishers say more than a hundred from anywhere. The trend matters more than the absolute number: citations growing steadily from credible sources is the signal that other people now treat you as a reference.
Branded demand
Finally, are people searching for your name? Branded demand is the hardest signal to fake and often the last to move, which is exactly why it is so honest. When people start seeking you out by name, the niche has begun to recognise you as an entity, the deeper signal we explore in brand signals and search.
Reading the signals together
The value is in the combination. A few patterns recur often enough to be worth naming.
- Coverage up, rankings up, citations up, brand up: you are genuinely gaining authority. Keep doing what you are doing.
- Coverage up but rankings flat: you are publishing but not earning. The content may be thin, or you lack the citations to make it rank. Deepen and promote rather than publish more.
- Rankings up but brand flat: you are ranking on page strength alone and have not yet become an entity. Invest in citable assets and consistent naming.
- Everything flat after growth: you have plateaued, often because the best pages have gone stale. This is the cue to refresh rather than to expand.
That last pattern is why measurement and maintenance belong together. A slow ranking slide on your strongest pages is usually the first sign that authority is decaying, and catching it early lets you act before the position is lost, the discipline we cover in our work on sustained authority.
The traps in authority measurement
Measurement goes wrong in predictable ways. The most common is vanity: tracking total traffic, which can rise for reasons unrelated to authority, or total page count, which rewards publishing volume regardless of quality. Neither tells you whether you own a topic.
The second trap is over reliance on third party authority scores. They have their uses as rough comparisons, but they are modelled approximations, not ground truth, and optimising for them directly is a fast route to wasted effort. We use them as a sanity check, never as a target.
The third trap is measuring too often. Authority moves slowly, and daily fluctuations are mostly noise. We read the trend over months, because that is the timescale on which the signals actually mean something. Checking daily produces anxiety, not insight.
How we use the read in practice
The point of measuring is to decide what to do next, not to produce a dashboard. We use the Coverage and Citation read to choose between three actions: deepen the pages we already have, expand into parts of the topic we do not yet cover, or refresh the pages whose signals are slipping. The signals tell us which lever to pull. Coverage gaps say expand. Flat rankings on covered queries say deepen or promote. Slipping rankings on strong pages say refresh.
Used this way, measurement stops being reporting and becomes steering. That is the only reason to do it. The thinking behind which topics are even worth measuring our share of sits in our wider building thesis, because the first measurement decision is always which topic to commit to in the first place.
The honest bottom line
Topical authority cannot be reduced to a number, and anyone who sells you one is selling a simplification. What you can do is watch coverage, ranking spread, citations, and branded demand together, read their combined direction over months, and let that direction guide where you put the next unit of effort. Done patiently, that read is more reliable than any single score, because it measures the thing itself rather than a proxy for it.
Kings Hospitality Group judges topical authority by a Coverage and Citation read: the share of a mapped topic where we rank on page one, set against the number of independent sites citing us. We track the trend of both together rather than any single published figure.
Common questions
Is there a single topical authority score I can track?
No reliable one exists. Third party scores are rough approximations, not ground truth. Measure authority by watching coverage, ranking spread, citations, and branded demand together over months, and read their combined direction rather than any single figure.
How often should I measure topical authority?
Monthly is usually right. Authority moves slowly, so daily checks are mostly noise. Reading the trend over months tells you whether the signals are genuinely moving, and gives you time to act before a slipping page loses its position.